Isoenergetic model for optical downconversion and error-specific limits of the parametric approximation

D. B. Horoshko, V. S. Shchesnovich · 2025

Read Isoenergetic model for optical downconversion and error-specific limits of the parametric approximation by D. B. Horoshko, V. S. Shchesnovich (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-18828.txt.

Introduction

Splitting one photon into a pair of lower-energy photons is a workhorse tool of quantum optics. D Horoshko and V Shchesnovich refine the math of that process in a 2025 paper, going beyond the standard shortcut that treats the incoming beam as fixed.

Their improved model lets the pump beam drain and tangle with the light it creates. It can compute field averages far past the point where the old shortcut fails. It also sets precise error limits for the old model: for a given pump strength, it names the most squeezing the shortcut can describe within a set error. Recent boson sampling experiments sit safely inside the valid zone, but the margin shrinks fast as squeezing grows.

Catalog record

CREATORD. B. Horoshko, V. S. Shchesnovich
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE77 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-18828

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Isoenergetic model for optical downconversion and error-specific limits of the parametric approximation D. B. Horoshko, V. S. Shchesnovich Abstract Optical downconversion is widely used for generating photon pairs, squeezed and entangled states of light, making it an indispensable tool in quantum optics and quantum information. In the regime where the pump is much stronger than the generated field, the standard parametric approximation treats the pump amplitude as a fixed parameter of the model. This approximation has a limited domain of validity since it assumes a non-depleted and non-entangled pump. By finding an approximate solution to the Schrödinger equation of the downconversion process, we obtain an improved analytical model beyond the parametric one, which accounts for pump depletion and pump-signal entanglement. The new model is advantageous, first, because it allows one to compute averages of field operators far beyond the domain of validity of the parametric approximation, and second, because it allows one to establish error-specific limits of the latter domain. For a given pump amplitude, we find a maximum squeezing parameter, up to which the approximation remains valid within a specified acceptable error. Our results confirm that recent experiments on Gaussian boson sampling, with a squeezing parameter of r ≈ 1.8 r\approx 1.8 and a coherent pump amplitude of α ≈ 2 ⋅ 10 6 \alpha\approx 2\cdot 10^{6} , can still be accurately described by the standard parametric approximation. However, we observe a sharp decline in validity as the squeezing parameter increases. For pump amplitudes of α ≈ 2 ⋅ 10 6 \alpha\approx 2\cdot 10^{6} , the parametric approximation breaks down when the squeezing parameter exceeds r ≈ 4.5 r\approx 4.5 , whereas the new approximation remains valid up to r ≈ 6 r\approx 6 with an acceptable error of 1%. I Introduction Optical parametric downconversion [ 1 , 2 ] is a versatile tool for classical optical applications such as frequency conversion and parametric amplification, as well as an essential source of entangled and squeezed states of light for photonic quantum technologies [ 3 ] . In most applications, the pump is treated as an undepleted classical wave that modulates the coupling parameter of the subharmonic fields, an approach known as the parametric approximation [ 4 ] . This approximation is sufficient in the low-gain regime, where single photon pairs are generated [ 5 , 6 ] , and in the high-gain regime with a moderate degree of squeezing [ 7 , 8 , 9 ] . However, many quantum information tasks, such as quantum-enhanced interferometry [ 10 , 11 ] , cluster state quantum computation [ 12 , 13 ] , and demonstrations of quantum advantage through Gaussian boson sampling [ 14 , 15 , 16 ] , require a high degree of squeezing. Progress in this area is very fast, and current experiments approach the quantum advantage regime [ 17 , 18 , 19 , 20 ] . Although higher squeezing levels can be achieved by increasing the pump power and using longer nonlinear crystals, this also results in a larger fraction of the pump energy being transferred to the signal field, thereby challenging the validity of the parametric approximation. The limits of the parametric approximation were first established in the seminal work of Hillery and Zubairy [ 21 ] , who analyzed generation of a single-mode signal field under the interaction Hamiltonian {H}=i\hbar\kappa\left({a}{b}^{{\dagger}2}-{a}^{\dagger}{b}^{2}\right), (1) where a and b are the photon annihilation operators for the pump and signal modes, respectively, and \kappa is the coupling constant proportional to the nonlinear susceptibility of the crystal. In the parametric approximation, the pump is assumed to be in a coherent state with amplitude \alpha . This allows one to make the substitution a\to\alpha e^{-i\omega_{p}t} , where \omega_{p} is the pump frequency, leading to a parametric Hamiltonian that transforms the initial vacuum state of the signal mode into the squeezed state |r\rangle with the squeezing parameter r . Using a path-integral approach, Hillery and Zubairy computed certain correlation functions for the signal-mode operators up to the first order in a perturbative expansion, where the zeroth-order term corresponded to the parametric approximation. They also identified the conditions under which this approximation remains valid: specifically, when each component of the vector (1/\alpha,r/2\alpha,re^{2r}/2\alpha,e^{2r}/\alpha) is much smaller than 1. As noted by the authors, these conditions are somewhat redundant and can be rewritten more concisely as \max(2,r)e^{2r}/2\alpha\ll 1 . While this limitation has been sufficient for quantum optical experiments over the past 40 years, the increasing demands of modern applications now call for a more refined limitation. Many modern quantum technologies require approximation limits to be error-specific. For example, a Gaussian boson sampler [ 15 , 16 ] enables observation of the statistical frequency F_{k} of the k th outcome in the distribution of a certain number of indistinguishable bosons across a given set of modes. This frequency asymptotically approaches the probability P_{k} predicted by the quantum model. For sufficiently large numbers of bosons and modes, the experimental observation of F_{k} happens much faster than the calculation of P_{k} on a classical supercomputer, which constitutes the principle of quantum advantage. However, for a finite sample size, F_{k} never coincides exactly with P_{k} , and the experimental result is accepted as a solution if it lies in the \epsilon -vicinity of the exact solution, where the distance can be measured, e.g., by the Euclidean norm \epsilon=\sqrt{\sum_{k}(F_{k}-P_{k})^{2}} or some other measure. In addition, boson samplers are subject to losses and indistinguishability degradation, e.g., due to the multi-mode structure of squeezed light pulses [ 22 , 23 , 24 ] . Specifically, when the overall transmission \eta of the boson sampler drops below a threshold value \eta_{\infty}(r) , the outcome probabilities can be efficiently calculated using a classical algorithm, eliminating the quantum advantage [ 22 ] . The threshold transmission \eta_{\infty}(r) can be reduced below the actual transmission \eta by increasing the squeezing parameter r , making the exact classical simulation infeasible. However, an approximate simulation would still be possible: There would exist a fast classical algorithm to calculate a probability distribution \tilde{P}_{k} that lies within the \epsilon^{\prime} -vicinity of the exact solution P_{k} . For the quantum advantage to hold, it is necessary that \epsilon\ll\epsilon^{\prime} , meaning the boson sampler produces results significantly closer to the exact solution than any classical simulation. Increasing the squeezing parameter would eventually approach the limits of applicability of the parametric approximation, which would result in the growth of \epsilon , because the quantum model (in the parametric approximation) would become less precise. Consequently, approximation limits must be formulated in an error-specific manner as V(r,\alpha,\epsilon)<1 , where V(r,\alpha,\epsilon) is a real indicator function defining the validity region in the (r,\alpha) space for a given acceptable error \epsilon . The Hillery-Zubairy limit can be rewritten in this form using the indicator function V_{HZ}(r,\alpha,\epsilon)=\max(2,r)e^{2r}/2\alpha\epsilon . However, as we demonstrate, this result holds only for sufficiently large values of \epsilon (above 10^{-2} ), whereas modern boson samplers require much smaller acceptable error thresholds. In this paper, we introduce a new approach to the analytical calculation of the joint quantum state of pump and signal for the Hamiltonian given by Eq. ( 1 ). Unlike the parametric approximation, this solution is consistent with the field-energy

Read the complete text